1. Field of the Invention
The present invention relates to a language processing function measuring device for conducting experiments relating to language processing of subjects, and more particularly for automatically selecting words for presentation to a subject under preset conditions while minimizing errors in the measured data.
2. Description of Related Art
A language experiment (for example Japanese Patent Laid-Open Publication No. 2001-37733), such as a priming experiment where a first word (prime), which is regarded to as precedent stimulant, is presented to a subject, and after a certain period of time has passed, a second word (target), which is regarded to as subsequent stimulant, is presented, and the reaction time to the second word is measured, is conventionally conducted.
In this language experiment, it is necessary to prepare a list of words to be presented to the subject in advance, and conventionally, in the preparation of this list of words, words whose category and lexical properties match desired conditions are extracted mainly from materials, such as newspapers, magazines, books or dictionaries, by hand.
However, it takes time and efforts for this extraction task because full determination has to be made to each word from gigantic materials; in addition, inappropriate words have to be excluded according experience and propriety has to be confirmed with respect to a subject group, which is correspondent to target age, sex and educational background, using a questionnaire(s).
Further, since this extraction task is conducted by hand, even though words satisfy set conditions, the degree of satisfying the conditions is different per word, and as a result, it will adversely affect measurement data, and then, a defect not to be able to obtain reliable experimental results will also occur.
In addition, more than a certain number of words are required for the experiment. However, because there is a so-called trade-off relationship between the number of words to be extracted and the number and/or severity of set conditions, if the extraction task is conducted by hand, it is very difficult to find appropriate conditions for a short time and to obtain the desired number of words, and there is also a problem that it takes a long time for trial and error regarding which condition can be relaxed for the purpose of obtaining the desired number of words or how to set the conditions.